The major goal of this project is to determine whether PET scans ( 18F) fluorodeoxyglucose (PET-FDG) to measure metabolic rate for glucose and (11C)-PK 11195 to assess peripheral benzodiazepine binding sites can predict the outcome of newly diagnosed patients with gliomas. The predicative power of PET will be compared with three novel methods of assessing the growth kinetics of gliomas, including the bromodeoxyuridine labeling index, Ki-67 antibody binding, and c-myb proto-oncogene expression. In addition, we will assess the ability of gadolinium MRI to predict outcome. The results of PET imaging, growth kinetic indices, and MRI will be correlated with neuropathological grade and with two measures of outcome, time to progression and total survival. The results of this study may support the routine clinical application of one or more of these new techniques for assessment of patients with gliomas, permitting more rational choice of treatment. This might be especially important in patients with low-grade gliomas in whom initial treatment is controversial. Secondary goals include determining whether PET-FDG scans can improve the diagnostic accuracy of stereotactic biopsies and whether high metabolic rate corresponds to specific pathological characteristics. The results could lead to increased accuracy of pathological diagnosis of gliomas.